CN103595780A - Cloud computing resource scheduling method based on repeat removing - Google Patents

Cloud computing resource scheduling method based on repeat removing Download PDF

Info

Publication number
CN103595780A
CN103595780A CN201310553508.XA CN201310553508A CN103595780A CN 103595780 A CN103595780 A CN 103595780A CN 201310553508 A CN201310553508 A CN 201310553508A CN 103595780 A CN103595780 A CN 103595780A
Authority
CN
China
Prior art keywords
virtual machine
node
computing node
scheduling
semantic
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201310553508.XA
Other languages
Chinese (zh)
Other versions
CN103595780B (en
Inventor
付印金
倪桂强
姜劲松
端义峰
金凤林
胡谷雨
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
PLA University of Science and Technology
Original Assignee
PLA University of Science and Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by PLA University of Science and Technology filed Critical PLA University of Science and Technology
Priority to CN201310553508.XA priority Critical patent/CN103595780B/en
Publication of CN103595780A publication Critical patent/CN103595780A/en
Application granted granted Critical
Publication of CN103595780B publication Critical patent/CN103595780B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Abstract

The invention discloses a cloud computing resource scheduling method based on repeat removing. The method includes the following steps that compute nodes send resource scheduling requests to a management node; the management node evaluates the system load level of each compute node; whether available cloud computing resources can satisfy the scheduling requests or not is judged, if not, the scheduling process is ended, and if yes, the scheduling process is continued; the management node calculates the semantic similarity between a virtual machine to be scheduled and virtual machines running on the compute nodes; the management node weights the sum of the semantic similarity of the virtual machine to be scheduled and the virtual machines running on the corresponding compute nodes so as to find out a selectable target compute node; whether the selectable target compute node is unique or not is judged, if yes, the last step is executed, and if not, the steps are sequentially executed; a selectable node is randomly selected as the target compute node; the target compute node distributes the computing resources to the virtual machine to be scheduled so that the virtual machine to be scheduled can run. The method can simultaneously optimize the I/O performance of a cloud computing system, keep node cluster loads balanced, and improve the use ratio of the resources.

Description

Based on the heavy cloud computing resource scheduling method that disappears
Technical field
The invention belongs to virtual and cloud computing field, particularly a kind of based on the heavy cloud computing resource scheduling method that disappears.
Background technology
Cloud computing can utilize Intel Virtualization Technology that the computational resource of each physical node of data center is abstracted into many virtual machines with independent computing capability, and for user provides as required, take the calculation services that virtual machine is unit by Internet.Due to reasons such as the dynamic expansion of user's calculation services demand dynamic change, physical node maldistribution of the resources, cloud computing system and physical node replacements, make cloud computing resources dispatching technique become one of key technology of cloud computing, for keeping, cloud computing system is efficient, stable, vital effect is played in operation reliably.
In cloud computing system, virtual machine operates on physical computing node cluster, and by the file data of shared memory systems managing virtual machines.Virtual-machine data can be divided into operating system data, application data and user data three classes.Virtual-machine data is carried out storage administration in the mode of virtual machine image file.Because a large amount of virtual machines are used identical operating system and application software, user file even, causing mass data content is repetition.
The weight technology that disappears is as a kind of efficient data reduction technology, and more traditional data compression technique can be found more data redundancy, and it is widely used in each data management layer of cloud computing system.In the virtual-machine data stack of computing node, the similar virtual-machine data page of data content is disappeared heavily, not only can effectively improve the physical resource utilance of internal memory, local disk data buffer storage and the shared memory systems of computing node, can also, by reducing I/O operand and volume of transmitted data between computing node and shared memory systems, greatly improve the systematic function of computing node.
In cloud computing system, when user asks to create a new virtual machine, or existing virtual machine every a period of time after operation, do not restart, and when additions and deletions computing node causes virtual machine (vm) migration while safeguarding cloud computing system, all need to carry out cloud computing resources scheduling and move to virtual machine, with keep whole cloud computing system efficiently, reliably operation.Existing cloud computing resource scheduling method, as USENIX HotCloud ' 12 papers " Dynamic Virtual Machine Scheduling in Clouds for Architectural Shared Resources " (open day: 2012-06-12), the Chinese invention patent application large-scale virtual machine fast transferring decision-making technique of data center " a kind of facing cloud " (application number: 201310186581.8 open days: 2013-08-14) and Chinese invention patent application " a kind of dispatching method of virtual machine and system " (application number: 201310186581.8, open day: 2013-03-20), main taking into account system performance and load balance, and do not consider to carry out in computing node the data impact that heavily processing brings that disappears.
Visible, the problem that prior art exists is: how to optimize cloud computing system I/O performance, and when keeping the load balance of cloud computing node cluster, can also improve the resource utilization of cloud computing system.
Summary of the invention
The object of the present invention is to provide a kind of cloud computing resource scheduling method based on disappearing and weighing, when optimizing cloud computing system I/O performance and keeping the load balance of cloud computing node cluster, improve the resource utilization of cloud computing system.
The technical solution step that realizes the object of the invention is as follows: a kind of based on the heavy cloud computing resource scheduling method that disappears, cloud computing system comprises computing node cluster, management node and shared storage pool, comprises the steps:
10) cloud computing resources dispatch request sends: a computing node sends information such as treating the system resources in computation demand of scheduling virtual machine and the operating system of this virtual machine, software configuration and user to management node;
20) computing node load proficiency assessment: management node is assessed the system load level of each computing node in cluster;
30) whether sufficient judgement of resource: judge whether that enough available computational resources operations treat scheduling virtual machine, no, scheduling of resource process finishes; To continue subsequent step;
40) virtual machine semantic similarity calculates: quantification calculates this and treats the semantic similarity between scheduling virtual machine and each computing node operation virtual machine;
50) optional target computing node is selected: utilize semantic similarity that the inverse weight of each computing node load level value treats all operation virtual machines in scheduling virtual machine and respective nodes with, select maximum weighted semantic similarity and corresponding computing node as optional target computing node;
60) optional target computing node quantity judgement: if optional target computing node only has, jump to target computing node distributes calculation resources (80) step, not only one of optional target computing node, continues subsequent step;
70) target computing node is selected: in the set forming at a plurality of optional target computing nodes, choose at random the target computing node that scheduling virtual machine operation is treated in a conduct;
80) target computing node distributes calculation resources: distributes calculation resources is given and treated scheduling virtual machine operation on destination node.
Compared with prior art, significant advantage of the present invention is:
1, the utilance of system resource is high: based on virtual machine semantic analysis, the scheduling virtual machine that content is similar arrives same computing node, make redundant data concentrate on intra-node, after heavily processing by disappearing, greatly saved the space of computing node internal memory and local disk data buffer storage and used;
2, systematic function is good: based on disappearing, heavily process and Data cache technology, the required mass data page calling of virtual machine operation has all existed in internal memory and local disk data buffer storage, need to not repeat to read from shared storage pool.Meanwhile, those data pages of having write do not need again to write shared storage pool yet.This has greatly reduced the data I/O operation between cloud computing node cluster and backstage shared storage pool, thereby has improved virtual machine runnability;
3, system load balancing: instruct scheduling virtual machine with the system load of balance node according to the load level of each computing node, improved the extensibility of system.
Below in conjunction with accompanying drawing, the present invention is described in further detail.
Accompanying drawing explanation
Fig. 1 is cloud computing system configuration diagram.
Fig. 2 is based on the heavy data page membership credentials schematic diagram that disappears.
Fig. 3 is shared storage pool data layout schematic diagram.
Fig. 4 the present invention is based on the heavy cloud computing resource scheduling method flow chart that disappears.
Embodiment
Fig. 1 is a kind of based on the heavy cloud computing system configuration diagram that disappears, and comprises cloud computing node cluster, management node and three parts of shared storage pool.
1, cloud computing node cluster: each node is invented a plurality of virtual machines, virtual machine manager is in charge of the operation of these virtual machines, each node operation has the heavy engine that disappears to come the space of memory optimization and local disk data buffer storage to use, and has resource monitor Real-time Collection physical node system resource service condition.
2, management node: realize the core of this method, be responsible for analysis, search index and the similarity of virtual machine semanteme and calculate, computing node during computing node cluster load evaluation, and scheduling virtual machine is in addition selected.
3, shared storage pool: be responsible for depositing virtual machine image file and virtual machine semantic indexing.
Each physical node based on the heavy cloud computing node cluster that disappears is invented a plurality of virtual machines, for user provides calculation services, and the shared use by these virtual machines of the unified coordination of virtual machine manager to bottom physical resource.Disappear heavy engine by elimination of duplicate data, reduce the space use amount of data in the local internal memory of node and disk unit.Computing node this locality has data buffer storage and preserves the recently data of frequent access, and to optimize the I/O performance of virtual machine access, also having resource monitor provides real-time system load information for management node.
Fig. 2 is based on the heavy data page membership credentials schematic diagram that disappears.By node data, disappear after heavily processing, no matter be between each virtual machine, or virtual machine is inner, and the logical page (LPAGE) of identical content is all shared Same Physical page in internal memory and local disk data buffer storage.
Fig. 3 is shared storage pool data layout schematic diagram.Shared storage pool provides mobile sms service for virtual machine image file.It is divided into virtual machine semantic indexing district and virtual machine image file memory block.The semantic indexing item of virtual machine comprises: virtual machine ID, virtual machine semantic information and node ID.Virtual machine ID is a unique identification that has moved virtual machine.Operating system that virtual machine semantic information comprises virtual machine is semantic, application software is semantic and user semantic information.In the corresponding cloud computing node cluster of node ID, move certain node of this virtual machine.
Management node is responsible for dispatching the computational resource in cloud computing node cluster, managing virtual machines semantic indexing structure, set up virtual machine ID the mapping relations of corresponding virtual machine image file between each data segment of shared storage pool.No matter be that user asks to create a new virtual machine, still restart existing virtual machine, and additions and deletions cloud computing node is while causing moving virtual machine, all needs to carry out cloud computing resources by management node and dispatch to keep that whole system is efficient, stable, operation reliably.
Fig. 4 the present invention is based on the heavy cloud computing resource scheduling method flow chart that disappears.Based on the heavy cloud computing resource scheduling method that disappears, comprise following eight steps:
10) cloud computing resources dispatch request sends: a computing node sends information such as treating the computational resource requirements of scheduling virtual machine and the operating system of this virtual machine, software configuration and user to management node;
20) computing node load proficiency assessment: management node is assessed the system load level of each computing node in cluster;
Described computing node load proficiency assessment 20) step comprises:
21) system load information that each computing node obtains resource monitor sends to management node;
Described computing node load information obtains 21) in step, described computing node load information comprises the R of use amount of the system resources such as CPU ability, memory headroom, disk space and the network bandwidth c, R m, R d, R nwith total amount T c, T m, T d, T n.
22), according to the ratio of computing node sorts of systems resource use amount and total amount, choose maximum as the load level value of node
L i=max{R C/T C、R M/T M、R D/T D、R N/T N}，
In formula, N is cloud computing node clustered node number;
23) by step 22) assessment obtains the system resource load level value L of each computing node 1, L 2..., L n;
24), according to the use amount of each computing node system resource and total amount information, calculate the available quantity of each computing node system resource;
30) whether sufficient judgement of resource: judge whether that enough available computational resources operations treat scheduling virtual machine, no, scheduling of resource process finishes; To continue subsequent step;
40) virtual machine semantic similarity calculates: quantification calculates this and treats the semantic similarity between scheduling virtual machine and each computing node operation virtual machine;
Described virtual machine semantic similarity calculating 40) step comprises:
41) management node is from the cloud computing resources scheduling request information content of step 10), gathers and treats that the operating system of scheduling virtual machine is semantic, application software is semantic and the information such as user semantic;
Described operating system semantic information comprises OS Type OS twith the OS of version number v, described application software semanteme comprises the type App of various application nand the App of version number v, described user semantic comprises each user name User nwith its recent renewal time T ime u.
42) management node is inquired about the semantic information of having moved virtual machine in each computing node on virtual machine semantic indexing;
The virtual machine semantic indexing of depositing in management node inquiry shared storage pool, is used for than the semantic information for the treatment of between scheduling virtual machine and existing operation virtual machine.Virtual machine semantic indexing has been preserved the mapping relations of virtual machine and its semantic information and the mapping relations of virtual machine and computing node, as shown in table 1.
Table 1. virtual machine semantic indexing topology example
Figure BDA0000411068540000051
43) treat that the semantic similarity that scheduling virtual machine and each computing node have moved between virtual machine is given by the following formula:
D ( X , Y ) = 0 , if : x 1 ≠ x 2 ; 1 , if : x 1 = x 2 and y 1 ≠ y 2 ; 2 , if : x 1 = x 2 and y 1 = y 2 .
44) semantic similarity for the treatment of scheduling virtual machine and each computing node is calculated by following formula:
Dist ( VM 1 , VM 2 ) = D ( S 1 , S 2 ) × w S + Σ i = 1 p Σ j = 1 q D ( A li , A 2 j ) × w A + Σ i = 1 m Σ j = 1 n D ( U 1 i , U 2 j ) × w U ;
45), by above two-dimentional semantic similarity computing formula, on management node, quantification calculates this and treats to have moved the semantic similarity between virtual machine on scheduling virtual machine and each computing node;
In various above, bivector S=<OS t, OS v>, A=<App n, App v> and U=<User n, Time u>, represents respectively the semantic information of virtual machine internal operating system data, types of applications data and each user data, bivector set: VM={S, A 1..., A j, U 1..., U krepresent to have j application, a k user's virtual machine, Dist (VM 1, VM 2) represent two virtual machine VM 1={ S 1, A 11..., A 1p, U 11..., U 1mand VM 2={ S 2, A 21..., A 2q, U 21..., U 2nsemantic similarity, w s, w aand w ube respectively definable weights.
50) optional target computing node is selected: utilize semantic similarity that the inverse weight of each computing node load level value treats all operation virtual machines in scheduling virtual machine and respective nodes with, select maximum weighted semantic similarity and corresponding computing node as optional target computing node;
Described optional target computing node selection 50) step comprises:
51) that treats scheduling virtual machine and computing node has allly moved virtual machine semantic similarity and has been expressed from the next:
SUM ( VM r , Node i ) = &Sigma; t = 1 k D ( VM r , VM it ) ;
Here, VM rfor treating scheduling virtual machine, Node ifor operation has k virtual machine { VM i1, VM i2..., VM ikcomputing node, SUM (VM r, Node i) for treating scheduling virtual machine VM rwith computing node Node isemantic similarity between upper all operation virtual machines and.
52) by above semantic similarity and computing formula, on management node, quantification calculates semantic similarity and the SUM that treats all operation virtual machines on scheduling virtual machine and each computing node 1, SUM 2..., SUM n;
53) nucleus module of management node scheduling virtual machine device, utilizes step 23) in the inverse of each node load level value come weighting treat all operation virtual machines on scheduling virtual machine and this node semantic similarity and, obtain effective semantic similarity,
SUM 1/L 1、SUM 2/L 2、…、SUM N/L N；
54) select maximum weighted semantic similarity
SUM T/L T=max{SUM 1/L 1,SUM 2/L 2,…,SUM N/L N}
Corresponding computing node T moves and treats scheduling virtual machine as optional target computing node;
60) optional target computing node quantity judgement: if optional target computing node only has, jump to target computing node distributes calculation resources (80) step, if not only one of optional target computing node continues subsequent step;
70) target computing node is selected: in the set forming at a plurality of optional target computing nodes, choose at random the target computing node that scheduling virtual machine operation is treated in a conduct;
80) target computing node distributes calculation resources: distributes calculation resources is given and treated scheduling virtual machine operation on destination node.
Described target computing node distributes calculation resources 80) step comprises:
81) management node sends and treats that the required system resource information of scheduling virtual machine is to destination node, by destination node distributes calculation resources, moves to this virtual machine;
82) ID, semantic information and the destination node ID that treat scheduling virtual machine are added on the virtual machine semantic indexing in shared storage pool as new operation virtual machine index entry.

Claims (5)

1. based on the heavy cloud computing resource scheduling method that disappears, related cloud computing system comprises computing node cluster, management node and shared storage pool, it is characterized in that, comprises the steps:
10) cloud computing resources dispatch request sends: computing node sends information such as treating the computational resource requirements of scheduling virtual machine and the operating system of this virtual machine, software configuration and user to management node;
20) computing node load proficiency assessment: management node is assessed the system load level of each computing node in cluster;
30) whether sufficient judgement of resource: judge whether that enough available computational resources operations treat scheduling virtual machine, no, scheduling of resource process finishes; To continue subsequent step;
40) virtual machine semantic similarity calculates: quantification calculates this and treats the semantic similarity between scheduling virtual machine and each computing node operation virtual machine;
50) optional target computing node is selected: utilize semantic similarity that the inverse weight of each computing node load level value treats all operation virtual machines in scheduling virtual machine and respective nodes with, select maximum weighted semantic similarity and corresponding computing node as optional target computing node;
60) optional target computing node quantity judgement: if optional target computing node only has, jump to target computing node distributes calculation resources (80) step, not only, continue subsequent step;
70) target computing node is selected: in the set forming at a plurality of optional target computing nodes, choose at random the target computing node that scheduling virtual machine operation is treated in a conduct;
80) target computing node distributes calculation resources: distributes calculation resources is given and treated scheduling virtual machine operation on destination node.
2. cloud computing resource scheduling method according to claim 1, is characterized in that: described computing node load proficiency assessment (20) step comprises:
21) system load information that each computing node obtains resource monitor sends to management node, and described computing node load information comprises the R of use amount of the system resources such as CPU ability, memory headroom, disk space and the network bandwidth c, R m, R d, R nwith total amount T c, T m, T d, T n;
22), according to the ratio of computing node sorts of systems resource use amount and total amount, choose maximum as the load level value of node
L=max{R C/T C、R M/T M、R D/T D、R N/T N}，
In formula, N is cloud computing node clustered node number;
23) by step 22) assessment obtains the system resource load level value L of each computing node 1, L 2..., L n;
24), according to the use amount of each computing node system resource and total amount information, calculate the available quantity of each computing node system resource.
3. cloud computing resource scheduling method according to claim 1, is characterized in that: described virtual machine semantic similarity calculates (40) step and comprises:
31) management node is from cloud computing resources scheduling request information content, gathers and treats that the operating system of scheduling virtual machine is semantic, application software is semantic and the information such as user semantic, and operating system semantic information comprises OS Type OS twith the OS of version number v, application software semanteme comprises the type App of various application nand the App of version number v, user semantic comprises each user name User nwith its recent renewal time T ime u;
32) on the virtual machine semantic indexing of management node, inquire about the semantic information of having moved virtual machine in each computing node, virtual machine semantic indexing is kept in shared storage pool, has recorded mapping relations between virtual machine and its semantic information and the mapping relations between virtual machine and operation place computing node;
33) treat that the semantic similarity that scheduling virtual machine and each computing node have moved between virtual machine is given by the following formula:
D ( X , Y ) = 0 , if : x 1 &NotEqual; x 2 ; 1 , if : x 1 = x 2 and y 1 &NotEqual; y 2 ; 2 , if : x 1 = x 2 and y 1 = y 2 . ;
34) semantic similarity for the treatment of scheduling virtual machine and each computing node is calculated by following formula:
Dist ( VM 1 , VM 2 ) = D ( S 1 , S 2 ) &times; w S + &Sigma; i = 1 p &Sigma; j = 1 q S ( A 1 i , A 2 j ) &times; w A + &Sigma; i = 1 m &Sigma; j = 1 n D ( U 1 i , U 2 j ) &times; w U ;
35), by above two-dimentional semantic similarity computing formula, on management node, quantification calculates this and treats to have moved the semantic similarity between virtual machine on scheduling virtual machine and each computing node;
In various above, bivector S=<OS t, OS v>, A=<App n, App v> and U=<User n, Time u>, represents respectively the semantic information of virtual machine internal operating system data, types of applications data and each user data, bivector set: VM={S, A 1..., A j, U 1..., U krepresent to have j application, a k user's virtual machine, Dist (VM 1, VM 2) represent two virtual machine VM 1={ S 1, A 11..., A 1p, U 11..., U 1mand VM 2={ S 2, A 21..., A 2q, U 21..., U 2nsemantic similarity, w s, w aand w ube respectively definable weights.
4. cloud computing resource scheduling method according to claim 1, is characterized in that, described optional target computing node selects (50) step to comprise:
41) that treats scheduling virtual machine and computing node has allly moved virtual machine semantic similarity and has been expressed from the next:
SUM ( VM r , Node i ) = &Sigma; t = 1 k D ( VM r , VM it ) ;
Here, VM rfor treating scheduling virtual machine, Node ifor operation has k virtual machine { VM i1, VM i2..., VM ikcomputing node, SUM (VM r, Node i) for treating scheduling virtual machine VM rwith computing node Node isemantic similarity between upper all operation virtual machines and;
42) by above semantic similarity and computing formula, on management node, quantification calculates semantic similarity and the SUM that treats all operation virtual machines on scheduling virtual machine and corresponding computing node 1, SUM 2..., SUM n;
43) nucleus module of management node scheduling virtual machine device, utilize the inverse of each node load level value in step (23) come weighting treat all operation virtual machines on scheduling virtual machine and this node semantic similarity and, obtain effective semantic similarity,
SUM 1/L 1、SUM 2/L 2、…、SUM N/L N；
44) select maximum weighted semantic similarity
SUM T/L T=max{SUM 1/L 1,SUM 2/L 2,…,SUM N/L N}；
Corresponding computing node T moves and treats scheduling virtual machine as optional target computing node.
5. cloud computing resource scheduling method according to claim 1, is characterized in that, described target computing node distributes calculation resources (80) step comprises:
51) management node sends and treats that the required system resource information of scheduling virtual machine is to target computing node, by target computing node distributing system resource, moves to this virtual machine;
52) ID, semantic information and the target computing node ID that treat scheduling virtual machine are added on the virtual machine semantic indexing in shared storage pool as new operation virtual machine index entry.
CN201310553508.XA 2013-11-08 2013-11-08 Cloud computing resource scheduling method based on the weight that disappears Active CN103595780B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201310553508.XA CN103595780B (en) 2013-11-08 2013-11-08 Cloud computing resource scheduling method based on the weight that disappears

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310553508.XA CN103595780B (en) 2013-11-08 2013-11-08 Cloud computing resource scheduling method based on the weight that disappears

Publications (2)

Publication Number Publication Date
CN103595780A true CN103595780A (en) 2014-02-19
CN103595780B CN103595780B (en) 2017-01-04

Family

ID=50085760

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310553508.XA Active CN103595780B (en) 2013-11-08 2013-11-08 Cloud computing resource scheduling method based on the weight that disappears

Country Status (1)

Country Link
CN (1) CN103595780B (en)

Cited By (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104391735A (en) * 2014-11-14 2015-03-04 深信服网络科技（深圳）有限公司 Method and system for scheduling virtual machines in virtualized all-in-one machine cluster
WO2016045489A1 (en) * 2014-09-22 2016-03-31 中国银联股份有限公司 System and method for load estimation of virtual machines in a cloud environment and serving node
CN105511952A (en) * 2014-09-22 2016-04-20 苏宁云商集团股份有限公司 Resource self-migration method and system based on cloud computing platform
WO2016107135A1 (en) * 2014-12-30 2016-07-07 中兴通讯股份有限公司 Memory management method and system, and computer storage medium
CN106170767A (en) * 2014-12-18 2016-11-30 华为技术有限公司 A kind of apparatus and method determining that virtual machine quantity adjusts operation
CN106227601A (en) * 2016-07-26 2016-12-14 东软集团股份有限公司 Method and apparatus for the resource distribution of virtual platform
CN106339254A (en) * 2015-07-15 2017-01-18 中兴通讯股份有限公司 Quick starting method and device, and management node for virtual machine
CN106462471A (en) * 2014-06-30 2017-02-22 微软技术许可有限责任公司 Opportunistically connecting private computational resources to external services
CN106471773A (en) * 2014-06-30 2017-03-01 微软技术许可有限责任公司 The distribution of integrated form global resource and load balance
CN107589980A (en) * 2017-08-01 2018-01-16 佛山市深研信息技术有限公司 A kind of dispatching method of cloud computing resources
CN109194703A (en) * 2018-06-29 2019-01-11 平安科技（深圳）有限公司 Processing method, electronic device and the medium of cloud platform inter-host communication load
WO2020199467A1 (en) * 2019-04-04 2020-10-08 平安科技（深圳）有限公司 Method for controlling communication load between network hosts, electronic device, and readable storage medium
CN113778668A (en) * 2021-08-18 2021-12-10 中国光大银行股份有限公司 Virtual resource deployment method, system and computer readable medium
CN115242662A (en) * 2022-09-22 2022-10-25 音信云(武汉)信息技术有限公司 Data resource allocation method and device based on cloud computing
WO2024007171A1 (en) * 2022-07-05 2024-01-11 北京小米移动软件有限公司 Computing power load balancing method and apparatuses

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2011110026A1 (en) * 2010-10-29 2011-09-15 华为技术有限公司 Method and apparatus for realizing load balance of resources in data center
CN102739798A (en) * 2012-07-05 2012-10-17 成都国腾实业集团有限公司 Cloud platform resource scheduling method with network sensing function
CN103258008A (en) * 2013-04-16 2013-08-21 武汉理工大学 Multilevel service resource discovery method and system based on user situations
CN103338228A (en) * 2013-05-30 2013-10-02 江苏大学 Cloud calculating load balancing scheduling algorithm based on double-weighted least-connection algorithm

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2011110026A1 (en) * 2010-10-29 2011-09-15 华为技术有限公司 Method and apparatus for realizing load balance of resources in data center
CN102739798A (en) * 2012-07-05 2012-10-17 成都国腾实业集团有限公司 Cloud platform resource scheduling method with network sensing function
CN103258008A (en) * 2013-04-16 2013-08-21 武汉理工大学 Multilevel service resource discovery method and system based on user situations
CN103338228A (en) * 2013-05-30 2013-10-02 江苏大学 Cloud calculating load balancing scheduling algorithm based on double-weighted least-connection algorithm

Cited By (25)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106462471A (en) * 2014-06-30 2017-02-22 微软技术许可有限责任公司 Opportunistically connecting private computational resources to external services
CN106471773A (en) * 2014-06-30 2017-03-01 微软技术许可有限责任公司 The distribution of integrated form global resource and load balance
CN105511952A (en) * 2014-09-22 2016-04-20 苏宁云商集团股份有限公司 Resource self-migration method and system based on cloud computing platform
CN105511953B (en) * 2014-09-22 2019-04-05 中国银联股份有限公司 Virtual machine load evaluation system, method and service node under cloud environment
CN105511952B (en) * 2014-09-22 2020-02-04 苏宁云计算有限公司 Resource self-migration method and system based on cloud computing platform
CN105511953A (en) * 2014-09-22 2016-04-20 中国银联股份有限公司 System and method for evaluating load of virtual machine under cloud environment, and service node
WO2016045489A1 (en) * 2014-09-22 2016-03-31 中国银联股份有限公司 System and method for load estimation of virtual machines in a cloud environment and serving node
US10528378B2 (en) 2014-09-22 2020-01-07 China Unionpay Co., Ltd. System and method for load estimation of virtual machines in a cloud environment and serving node
CN104391735A (en) * 2014-11-14 2015-03-04 深信服网络科技（深圳）有限公司 Method and system for scheduling virtual machines in virtualized all-in-one machine cluster
CN104391735B (en) * 2014-11-14 2018-11-06 深信服网络科技（深圳）有限公司 Virtualize dispatching method of virtual machine and system in all-in-one machine cluster
CN106170767A (en) * 2014-12-18 2016-11-30 华为技术有限公司 A kind of apparatus and method determining that virtual machine quantity adjusts operation
CN106170767B (en) * 2014-12-18 2020-05-15 上海朋邦实业有限公司 Device and method for determining number adjustment operation of virtual machines
WO2016107135A1 (en) * 2014-12-30 2016-07-07 中兴通讯股份有限公司 Memory management method and system, and computer storage medium
CN106339254B (en) * 2015-07-15 2020-06-12 中兴通讯股份有限公司 Method and device for quickly starting virtual machine and management node
WO2017008584A1 (en) * 2015-07-15 2017-01-19 中兴通讯股份有限公司 Virtual machine starting method and device, and management node
CN106339254A (en) * 2015-07-15 2017-01-18 中兴通讯股份有限公司 Quick starting method and device, and management node for virtual machine
CN106227601A (en) * 2016-07-26 2016-12-14 东软集团股份有限公司 Method and apparatus for the resource distribution of virtual platform
CN107589980A (en) * 2017-08-01 2018-01-16 佛山市深研信息技术有限公司 A kind of dispatching method of cloud computing resources
CN109194703A (en) * 2018-06-29 2019-01-11 平安科技（深圳）有限公司 Processing method, electronic device and the medium of cloud platform inter-host communication load
WO2020199467A1 (en) * 2019-04-04 2020-10-08 平安科技（深圳）有限公司 Method for controlling communication load between network hosts, electronic device, and readable storage medium
CN113778668A (en) * 2021-08-18 2021-12-10 中国光大银行股份有限公司 Virtual resource deployment method, system and computer readable medium
CN113778668B (en) * 2021-08-18 2023-01-17 中国光大银行股份有限公司 Virtual resource deployment method, system and computer readable medium
WO2024007171A1 (en) * 2022-07-05 2024-01-11 北京小米移动软件有限公司 Computing power load balancing method and apparatuses
CN115242662A (en) * 2022-09-22 2022-10-25 音信云(武汉)信息技术有限公司 Data resource allocation method and device based on cloud computing
CN115242662B (en) * 2022-09-22 2023-02-17 音信云(武汉)信息技术有限公司 Data resource allocation method and device based on cloud computing

Also Published As

Publication number Publication date
CN103595780B (en) 2017-01-04

Similar Documents

Publication Publication Date Title
CN103595780A (en) Cloud computing resource scheduling method based on repeat removing
Kalia et al. Analysis of hadoop MapReduce scheduling in heterogeneous environment
Tian et al. Latency critical big data computing in finance
CN103617087A (en) MapReduce optimizing method suitable for iterative computations
Liu et al. SP-Partitioner: A novel partition method to handle intermediate data skew in spark streaming
CN107436813A (en) A kind of method and system of meta data server dynamic load leveling
CN103297499A (en) Scheduling method and system based on cloud platform
CN106293944A (en) System and optimization method is accessed based on nonuniformity I/O under virtualization multi-core environment
CN103198097A (en) Massive geoscientific data parallel processing method based on distributed file system
CN103491155A (en) Cloud computing method and system for achieving mobile computing and obtaining mobile data
Li et al. Data locality optimization based on data migration and hotspots prediction in geo-distributed cloud environment
CN111381928A (en) Virtual machine migration method, cloud computing management platform and storage medium
Irandoost et al. Mapreduce data skewness handling: a systematic literature review
US10481800B1 (en) Network data management protocol redirector
CN104158875A (en) Method and system for sharing and reducing tasks of data center server
CN101776946B (en) Method for controlling power consumption on basis of object-based storage system
CN107589980A (en) A kind of dispatching method of cloud computing resources
Liu et al. An efficient job scheduling for MapReduce clusters
Vadivel et al. Cauchy particle swarm optimization (cpso) based migrations of tasks in a virtual machine
CN105306547A (en) Data placing and node scheduling method for increasing energy efficiency of cloud computing system
CN106027685A (en) Peak access method based on cloud computation system
Zhao et al. A holistic cross-layer optimization approach for mitigating stragglers in in-memory data processing
CN104699520B (en) A kind of power-economizing method based on virtual machine (vm) migration scheduling
JP2012504814A (en) System and method for dynamic management of data centric search
Jolfaei et al. Improvement of job scheduling and tow level data replication strategies in data grid

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant